392 Records of the Indian Museum. (Vora 
Although the free gemmules do not differ so much from the 
fixed ones as is the case in C. ultima var. spinosa, the differences 
are of the same nature. In both kinds of gemmules the shape 
(apart from the external case) is almost spherical and there is a 
single aperture with a straight foraminal tubule, but the free gem- 
mule is slightly more flask-shaped than the other and has a longer 
and more tapering tubule. The diameter of both varies somewhat, 
but in the case of the free gemmule it is on an average about 0°5 
mm. and in that of the fixed gemmule about 06 mm. The free 
gemmule has a well-developed pneumatic coat outside its proper 
spicules and, outside this coat, is enclosed in a hollow sphere of un- 
usually small and often ill-formed megascleres mixed with spicules 
like its own and held together by a chitinous membrane. The 
microscleres are very little longer or more slender than is the case 
in the fixed gemmules. 
II.—GEOGRAPHICAL, 
In the introduction to my volume on the Freshwater Sponges, 
etc.,in the Fauna of British India J laid great stress on the African 
affinities of the lower invertebrates that inhabit the streams, lakes 
and pools of the Western Ghats. So far as the sponges are con- 
cerned the chief foundation for this view lies in the strong represen- 
tation of the genus Corvospongilla and the subgenus Stvatospongilla 
of the genus Spongilla. Recent investigations have on the whole 
given support to my belief but have shown that the African ele- 
ment is more widely distributed than was at first realized. The 
discovery of a freshwater medusa of the genus Limnocnida in 
the Western Ghats is evidence in favour of African affinities, 
and so also is the fact that three additional forms of Corvospon- 
gilla have now been added to the known fauna of the range, as 
well as anew species of Stratosbongilla; but on the other hand 
two species of Corvospongilla have been found in India east of the 
Ghats and the range of Spongilla (Stratospongilla) bombayensis, 
formerly believed to be peculiar to Bombay and Natal, is now 
known to extend into the W. Himalayas in the north and to the 
Mysore plateau in the south, while there is strong evidence that 
Limnocnida indica only occurs in those streams which run east- 
wards from the Ghats. 
Of the genus Corvospongilla three species, C. loricata (Welt- 
ner), C. bohm (Hilgendorf) and C. zambestana (Kirkpatrick), and 
possibly a fourth only recognized from isolated spicules are known 
to occur in Africa. The first, which is the type of the genus, is 
from an unknown locality in that continent, while the other three 
occur in Central Africa. Of the four Indian and Burmese species, 
one (C. burmanica) is very closely allied to C. loricata, while 
another (C. dapidosa) is perhaps no more than a local race of the 
African C. zambesiana, the gemmules of which are not known. 
The two sponges apparently differ only in the structure of their 
skeleton. All the Indian forms, with the exception of the typical 
